SAMPE Journal - May/June 2016 - 12

Feature Article

Figure 4. Demonstrator after lay-up showing the sensor
locations and IML ply-cuts (dart locations indicated by
dashed lines).

The formed inserts are shown in Figure 1. No
cell-buckling or node splitting was observed. Good
conformity was observed except for 2 mm of springback was observed for the double curvature insert.
This was noted during the IML skin lay-up; one
laminator pushed the core against the OML skin while
the other draped the prepreg down the chamfer.
The part was built with a [0]4 OML and a [0]4
IML skin, where the 0° direction was parallel to the
1400 mm edge in Figure 1. No CAD models were
available for this mould, therefore all ply cutting and
draping decisions were made without input from
ply placement software. A vacuum bag was used to
trace the approximate ply cutting template. All plies
were laid-up from the double-curvature region of the
mould, moving towards the flat region. Significant
material deformation in the flat-bottom depression
of the double-curvature region (Figure 1C) resulted
in the plies straying from the edge breathing, towards
the mould edge. The edge breathing was considered
essential for good gas evacuation, therefore some ply
cutting was required in both the OML and IML skins
to mate the plies onto the mould without wrinkling.

Before laying-up the first ply, the mould was preheated to 50°C before being moved into the clean
room. The surfacing film was placed onto the warm
mould without tearing. The surfacing film roll width
was insufficient to cover the mould, therefore two
pieces were used, with an overlap in the middle of
the flat region. The pre-heated mould increased
the unsupported surfacing film's tack, increasing
handling/placement difficulty. The first OML ply was
placed on the mould, followed by the edge breathing
and the final three OML plies. The OML skin was debulked for 2 h at ambient temperature and 1 h at 50°C
to consolidate the plies. The vacuum level was 120
mbar in both de-bulking stages.
Once the tool had cooled, the film adhesive and
honeycomb core were located on the part. The lay-up
was de-bulked at ambient temperature overnight inorder to consolidate the OML skin and core assembly
before applying the IML plies.
The IML skin was more challenging to lay-up because
the prepreg needed to conform to the honeycomb core
inserts. Even after the extended de-bulk, draping the
IML plies over the compliant core was difficult. As a
result, additional ply cutting was required for the IML
plies. Darts were introduced to avoid major wrinkling
of the plies. The location of the ply-cuts and darts for
each IML ply is shown in Figure 4. The first IML ply was
the most difficult to drape, and required three darts.
The second ply was easier to lay-up, requiring only one
dart. The final two IML plies were more difficult than
the second ply, requiring two darts each. An ambient
temperature de-bulk was performed after each ply. The
vacuum level was 120 mbar, and the hold time was 30,
90, 30, and 20 minutes for IML ply 1, 2, 3, and 4.
Process Monitoring Miniature Pressure Sensors
Miniature pressure sensors were used to measure the
honeycomb core pressure though the IML skin lay-up,
during the pre-cure vacuum hold, and during processing.
A patented miniature sensor technology from Convergent
Manufacturing Technologies21 was used to measure the
pressure inside the honeycomb core.
A piezoresistive silicon micromachined pressure
sensor from Measurement Specialties (model
number: MS5407-AM) was used in this work, and is
shown in Figure 5A. The sensor required four wires to
be soldered to the back side of the carrier to complete
the Wheatstone bridge. Enamel coated copper magnet
wire was used to avoid introducing leaks along the
wiring. A K-type thermocouple was embedded
alongside the pressure sensor for temperature
compensation.
SAMPE Journal, Volume 52, No. 3, May/June 2016

http://www.mikrosam.com

Table of Contents for the Digital Edition of SAMPE Journal - May/June 2016